1184 results about "Web technology" patented technology

The present invention discloses an artificial intelligence-based adaptive big data storage and retrieval optimization system and method designed to intelligently manage and optimize large-scale distributed data environments. The system integrates data acquisition, distributed storage, metadata processing, adaptive learning, and retrieval optimization units configured to work collaboratively for continuous self-optimization. The invention employs deep reinforcement learning and predictive neural network techniques to dynamically analyze system telemetry, workload behavior, and data access patterns in real time, enabling proactive adjustment of data placement, caching, replication, and compression parameters across distributed nodes. The metadata processing framework utilizes graph-based dependency modeling to maintain semantic and contextual relationships among datasets, facilitating intelligent and context-aware data retrieval. The retrieval optimization unit interprets user queries semantically and computes the optimal retrieval route using latency prediction models and dynamic routing techniques.

The invention relates to a power edge cloud cooperative management method based on a swan OS, and the method comprises the following steps: S1, deploying intelligent edge nodes, setting an embedded AI acceleration unit, operating the swan OS on all the intelligent edge nodes, and achieving the seamless cooperation of equipment through the distributed characteristics of the swan OS; and S2, realizing self-organizing connection of edge devices by adopting a Mesh network technology, designing a direct communication path between the edge devices, and realizing rapid processing of local problems. S3, aiming at the real-time requirement of a power system, optimizing an MQTT / CoAP protocol stack, improving the efficiency and reliability of message transmission, and allocating special network resources for different types of power applications by utilizing the 5G network slice characteristics, S4, constructing a security architecture based on a zero-trust principle, enhancing the equipment access control and data transmission security, and S5, establishing a security architecture based on a zero-trust principle. And identity authentication and data integrity protection are performed by using a block chain technology, so that data tampering and counterfeiting are prevented. According to the method, the local distributed decision-making capability and the global cooperation capability are effectively improved.

The invention relates to the technical field of wireless communication networks, in particular to a vehicle ad hoc network based on intelligent networked vehicle WiFi and a V2V interaction method, a vehicle motion state and network load data are acquired through a vehicle-mounted sensor, AP / STA modes are dynamically switched by combining a deep Q network (DQN) model and an epsilon-greedy strategy, and an adaptive topological structure is constructed. And expanding OLSR protocol message embedding path delay fields, calculating and optimizing multi-hop routing selection in combination with pheromone gradient, and realizing message priority classification and differential privacy protection through a federated learning model. And dynamically adjusting an AP / STA proportion threshold based on a particle swarm algorithm, and triggering node role reconfiguration to balance network load. According to the method, the existing WiFi hardware is reused, the deployment cost is reduced, cross-layer routing optimization and a distributed decision mechanism are combined, and the network real-time performance, the expansibility and the group cooperation capability in a dynamic scene are improved.

The invention relates to the technical field of software defined networks, in particular to an SRv6-based SDN (Software Defined Network) data center load balancing method and system, and the method comprises the steps: obtaining node real-time load and SRv6 routing available resource information through an SDN controller, constructing a load balancing model, and recognizing an overload node and a congestion link; generating a flow scheduling path in combination with segmented routing characteristics; in combination with network state information, monitoring the load condition in real time when the node is unbalanced or the link packet loss rate is gt; when 5%, a load balancing adjustment mechanism is triggered, unbalance types are distinguished, a related flow set is determined, optimization is performed by using a DQN algorithm, and a flow scheduling strategy is generated and dynamically adjusted by taking throughput maximization and unbalance degree minimization as targets; and issuing the scheduling strategy to the data center network equipment, updating the flow table through the Netconf protocol, and redistributing the flow. Therefore, the problems of low load balancing efficiency, insufficient dynamic adaptability, low resource utilization rate and the like in the prior art are solved.

The invention discloses a high-dynamic mobile ad hoc network dynamic time slot allocation method and device, equipment and a storage medium, and belongs to the technical field of wireless communication and networks. The device is deployed between an upper-layer application module and a bottom-layer communication module of a communication node, and comprises a topology sensing module, a time frame construction module, an idle time slot distribution module, a GNSS module and an information interaction module. According to the method, neighbor nodes are determined among nodes through interaction information, and node topology change indexes and link states are calculated; the coordination node performs idle time slot allocation according to a request sent by the node in the control time slot, and dynamically adjusts a sensing period according to the current network topology change; and the node sends data in the information time slot according to the time slot allocation table, and carries out information interaction and topology perception according to the perception period. According to the method, the packet loss rate in the link unstable environment is effectively reduced, the data transmission reliability is higher, the method can adapt to link and topology dynamic changes, and the network throughput and the resource utilization rate are improved.

The invention discloses a QoS guarantee method and system for a communication network, and relates to the technical field of communication networks, and the method comprises the steps: collecting and preprocessing network state data, and forming a standardized data set; dynamically classifying service types based on an improved random forest algorithm, and predicting a future QoS demand trend of each priority service in combination with an LSTM neural network; establishing a mapping model of QoS demands and resource parameters, converting predicted demands into allocable resource indexes, monitoring the resource utilization rate in real time, and setting an elastic reservation mechanism and conflict early warning; when early warning is triggered, selecting an optimal transmission link by adopting a multi-path collaborative algorithm, and implementing differentiated resource allocation according to service priorities; and a closed-loop feedback mechanism is triggered to dynamically adjust resource allocation by monitoring the deviation between the actual QoS and a predicted value in real time. The method has the advantages that through multi-dimensional perception, LSTM prediction, dynamic resource management and multi-path scheduling, QoS requirements of services with different priorities are accurately matched, and dynamic changes of the network are efficiently coped with.

The invention relates to the technical field of communication networks, in particular to a real-time network flow prediction and resource optimization distribution method and system based on a space-time multi-mode generation model. Multi-modal feature extraction and fusion are carried out; model prediction is generated; reinforcement learning scheduling optimization; self-supervision and feedback closed loop are realized; executing resource allocation; the method has the advantages that multiple data sources such as spatio-temporal data, user behaviors and application types are fused through the multi-modal large model, future network traffic distribution is generated through the generation model, and the accuracy and timeliness of traffic prediction are improved. The network resources of different areas and time periods are intelligently scheduled through the reinforcement learning model, real-time and automatic resource allocation is realized, and the network quality and user experience of high-demand areas are ensured.

The invention discloses a front-end task scheduling and event processing method and device, equipment and a storage medium, and relates to the field of distributed systems and network technologies, and the front-end task scheduling and event processing method comprises the steps of capturing events triggered by all sub-elements; processing the captured events into standard task objects in a unified format; adding the standard task objects into a scheduling queue according to a priority sequence, and performing de-duplication processing and task merging processing on all the standard task objects in the scheduling queue to obtain a cleaned scheduling queue; all the standard task objects are sorted again according to the priority in the multi-task priority queue; according to the priorities of the standard task objects in the multi-task priority queue, dynamically adjusting the resource allocation of each standard task object in the multi-task priority queue in combination with the current load of the system; and executing the standard task object allocated with the resources by combining an asynchronous execution mode with a user feedback mechanism. According to the method, the problems of task conflict and resource competition in high-concurrency interaction can be effectively solved.

The invention provides an Internet of Vehicles vulnerability management method, system, device and medium based on a block chain, and relates to the technical field of vehicle networks, the method can comprehensively identify potential safety risks of vehicles through a dynamic and static combined vulnerability detection mechanism, and generates a structured vulnerability report; hash abstract chaining evidence storage is performed on key data of reports and subsequent repair links by using a block chain, so that the whole process of vulnerability discovery, analysis and repair is ensured not to be tampered and traceable, and the authenticity and credibility of data are improved; the integrity of the report is verified at the cloud end, and an AI analysis engine is combined to associate a CVE database and threat intelligence, so that intelligent generation and decision support of a repair scheme are realized; the hash abstract on the chain of the OTA patch is verified at the vehicle end, so that the credibility of the patch source and the integrity of the content are guaranteed; and finally, through feedback of an installation result and secondary uplink archiving, complete closed-loop management from vulnerability discovery to repair verification is formed.

The invention relates to the technical field of networks, in particular to a remote sensing image semantic segmentation network and method, and the method comprises the steps: constructing a remote sensing image semantic segmentation model based on context anchor point residual errors and multi-scale aggregation, the network structure of the model is mainly composed of a context anchor point residual block CARB, a multi-scale attention aggregation module MSAA and a mixed residual path MRP. In the model training stage, a composite loss function of a cross entropy loss function and a Dice coefficient loss function is used for assisting model training; original input images of different sizes are input into the network, through a series of feature extraction and classification of the network, a segmentation task with clear boundaries is finally completed on a target, a complete segmentation effect picture is output, the network can extract richer feature information, fine spatial features and deep semantic information of the images are effectively captured, and the image segmentation efficiency is improved. And the target identification capability is further enhanced, so that a better segmentation effect is realized in a complex environment.

Techniques for computer science, data science, data analytics, computer software, algorithmic analysis, and networked technologies for sourcing, procurement, manufacturing, and supply chain management in small-to-medium manufacturing (“SMM”) industries that involve processes whereby implementing user input and extracted image and text from engineering drawings may be used to map manufacturing requirements to a subset of manufacturers via natural language agent programs. More specifically, natural language agent programs may be implemented to apply input and extracted image and text from engineering drawings to a large language model (“LLM”). An example method may include receiving a user input including data representing requirements to manufacture a physical structure, concatenating an engineering drawing summary data, a shoptype description, and estimated part size instruction data to combine the at least two of the engineering drawing summary data, the shoptype description, and the estimated part size instruction data to generate a subset (e.g., a list) of qualified manufacturers (e.g., SMMs).

The invention relates to the technical field of wireless communication networks, in particular to an eSIM global operator dynamic OTA switching system based on an AI strategy, and the system comprises an environment data collection module which collects the bandwidth, the packet loss rate, the network stability and the switching success rate at an equipment end, and generates local network state data; the behavior and tariff modeling module collects communication behavior and tariff information and generates user preference data; the AI strategy generation module outputs an OTA control instruction based on a graph neural network and an attention mechanism; the OTA strategy synchronization module issues and analyzes an instruction to generate an execution parameter packet; and the double-link switching execution module completes switching and rollback configuration. According to the invention, by fusing graph neural network strategy modeling, a terminal AI and OTA cooperation mechanism and a double-link redundancy switching process, intelligent, low-delay and high-reliability dynamic switching of an eSIM operator is realized.

The invention discloses a network security supervision system based on a network technology, and particularly relates to the technical field of data security, the network security supervision system comprises an intelligent detection terminal, an edge cooperation terminal, a cloud platform, a strategy management terminal, a management control terminal and a system database, the intelligent detection terminal is used for collecting network security data in real time and performing data preprocessing; the edge collaboration terminal is used for carrying out preliminary analysis on the integrated data and identifying common security threats; the cloud platform is used for carrying out multi-source data fusion, carrying out deep analysis and evaluating a network security situation; the strategy management terminal is used for formulating differentiated security strategies and dynamically adjusting the security strategies; the management control terminal is used for providing a visual operation interface for system management and decision making; and the system database is used for storing data for a long time and supporting a real-time retrieval function. By adopting a multi-layer distributed architecture and multi-source data fusion, the accuracy of security assessment is enhanced, and the efficiency of threat monitoring is improved.

The invention provides a network traffic scheduling optimization method based on deep learning. The method is applied to the technical field of communication networks, and comprises the following steps: S1, collecting flow data, link state, delay and packet loss rate indexes deployed at network nodes in real time, and constructing a sample database; s2, according to the sample database, constructing a traffic prediction model fusing a retrieval enhancement diffusion model and a mixed linear expert model to perform multi-modal traffic prediction; s3, constructing a state space of a reinforcement learning agent according to a real-time network state and the future traffic prediction result; s4, performing strategy optimization based on a multi-objective optimization mechanism; and S5, deploying the trained and optimized model to a network controller or an edge computing node to realize real-time sensing and dynamic scheduling control of network resources. According to the method, the traffic prediction precision is remarkably improved, the scheduling strategy is updated and optimized in time according to the network environment change, and the comprehensive performance of the network is improved.

The invention discloses a vehicle fault positioning and rescue scheduling platform and method based on GPS track acquisition, and relates to the field of vehicle fault positioning and rescue scheduling, and the method comprises the steps: uploading vehicle GPS and OBD data obtained by a vehicle-mounted terminal to a cloud database through employing a mobile network technology; according to the vehicle GPS and OBD data in the cloud database, the vehicle fault type is predicted and judged, and the position coordinate information of the vehicle fault is obtained; based on an expert scoring method, processing response levels of different vehicle fault types are set; accessing a high-precision digital map API (Application Program Interface), and marking schedulable rescue points and fixed GPS (Global Positioning System) coordinate data in a digital map; and establishing a vehicle fault rescue scheduling model, and carrying out comprehensive evaluation and scheduling on surrounding schedulable rescue points. The method has the advantages of effectively predicting the state and fault of the vehicle, efficiently responding to the processing of the fault of the vehicle, and preferentially scheduling and screening the schedulable rescue points.

The invention discloses a method and equipment for downloading a large file of a browser, belongs to the technical field of front-end Web, and is used for solving the problems that a large file downloading scheme in a traditional browser environment is low in memory use efficiency and data storage efficiency, insufficient in network interruption processing capability and lack of effective integration of a browser underlying technology, and the network interruption processing capability is poor. And offline caching and high-efficiency recovery are difficult to realize. The method comprises the steps of fragmenting a target file in a file downloading task according to a preset fragmentation rule, and creating a fragmentation downloading task for each fragmentation; dynamically distributing fragmented downloading tasks through a parallel task pool; after each fragment downloading task is completed, the downloaded fragment data and metadata thereof are stored through an IndexedDB database, and integrity verification of the fragment data is carried out; according to a fragment integrity verification result, re-downloading the current fragment; and after all fragment downloading tasks are executed, reading all fragment data according to a fragment sequence and splicing the fragment data into a complete file.

The invention relates to the technical field of communication networks, in particular to a dynamic resource allocation method and system for an industrial 5G network. The method comprises the following steps: acquiring industrial terminal data; identifying network congestion nodes based on industrial terminal data; detecting a link congestion fault according to the network congestion node, and determining a bandwidth utilization rate; allocating spectrum resources based on the bandwidth utilization rate; performing frequency spectrum slicing on the frequency spectrum resources, and determining a flow load amount; adjusting load balance according to the traffic load, and generating a load resource report; dynamically adjusting time slot allocation by using a load resource report, and determining a task priority; task scheduling simulation is executed according to the task priority, and if congestion occurs, a congestion snapshot is stored; reducing scheduling conflicts according to the congestion snapshots, releasing bandwidth resources, and determining available network slice capacity; power resources are dynamically adjusted based on available network slice capacity. The bandwidth utilization rate and the task scheduling efficiency of the industrial 5G network are improved based on the communication network technology.

The invention discloses a complex environment information acquisition method based on a sensor network, and relates to the technical field of sensor networks, and the method comprises the steps: building a complex environment portrait and a service level, and generating a scene configuration comprising an information freshness budget, a risk heat map and an alarm threshold; based on scene configuration, configuring mesh wireless networking, time synchronization parameters and candidate relays, and constructing observation configuration and quality rules; and generating a data stream and a real-time queue state according to a self-adaptive sampling strategy and an event mark, calculating an emergency degree and distributing a forwarding probability according to an information freshness budget by combining a candidate relay and a risk heat map, outputting uploaded data and a quality record, completing convergence and grading alarm according to an alarm threshold, and generating a disposal feedback and audit index. The state set, the uncertainty set, the link idle degree and the energy sufficiency are linked under the constraint of the information freshness budget, so that hot spot encryption, low-risk sparseness, reduction of energy consumption and conflict rate and standard observation time efficiency are realized.

The invention relates to the technical field of machine tool data processing, in particular to a machine tool abnormal operation data edge calculation method and system, and the method comprises the steps: deploying a sensor at a key part of a machine tool for collecting machine tool operation data and transmitting the machine tool operation data to an edge calculation device which has calculation capability and a local database; dynamically optimizing a data transmission path by adopting an SDN network technology in a data transmission process; after receiving the data, the edge computing device performs data preprocessing and abnormal data detection, stores the data and a detection result in a local database and regularly uploads the data and the detection result to a cloud server; and the cloud server analyzes the trend of abnormal operation of the machine tool by adopting a big data analysis algorithm, and feeds back optimization suggestions to the edge computing device according to an analysis result. According to the method, the data processing delay is reduced by combining the edge computing with artificial intelligence, the transmission path is optimized, the cooperation of the edge computing equipment and the cloud is realized, and the processing efficiency of the abnormal operation data of the machine tool is improved.

The invention relates to the technical field of security authentication and ubiquitous networks, and discloses a ubiquitous-network-oriented cloud-side collaborative multi-factor identity authentication method and system, and the method comprises the steps: issuing an identity certificate based on an elliptic curve and a PUF challenge-response library of a parameter initialization edge node through a cloud, and achieving the cloud-side bidirectional authentication; when the terminal is accessed, multi-factor authentication combining PUF challenge and a dynamic threshold value is adopted, so that the capability of resisting counterfeiting and replay attacks is improved; after the authentication is passed, the edge node generates a group public key locally and constructs a binary tree group key structure with a preset depth, and post-quantum safe and efficient group signature management is supported; when a newly added terminal joins, generating a secret key based on an elliptic curve algorithm and distributing a group signature private key share to realize lightweight member management; the integrity, confidentiality and traceability are realized while the transmission efficiency is guaranteed, and safe communication and dynamic revocation in an edge resource limited environment are effectively supported.

The invention relates to the technical field of communication networks, in particular to a resource allocation system and method in a satellite-ground fusion network, and the system comprises a satellite-ground topology and channel evolution predictor which is used for obtaining satellite ephemeris, user GNSS positioning and moving vectors, ground base station topology and a real-time CQI / RSRP measurement report; generating future link state information; the business QoE impact and switching cost modeling device is used for calculating transient interruption time of a switching action; generating switching cost information; the continuity risk quantification engine is used for generating continuity risk information containing the link maintaining risk probability, the instantaneous impact collapse probability and the link switching risk probability; the risk evasion type resource decision-making device is used for calculating maintenance link comprehensive cost and switching link comprehensive cost; calculating a dynamic switching damping threshold value; generating risk avoidance decision information; according to the invention, the high-frequency ping-pong switching near the edge state point is effectively inhibited, and the signaling overhead caused by invalid switching is reduced.

The invention relates to a method for constructing a cross-modal time sequence diagram model of a complex common disease network, and belongs to the technical field of common disease networks, and the method comprises the following steps: S1, dividing the diagnosis and treatment data of a patient into text data, image data and time sequence data classification; s2, processing text information by adopting a word embedding model, processing image information by adopting a convolutional neural network, and processing time sequence data by adopting a time sequence modeling module; then, a cross-modal multi-head attention CM-MHA module is adopted to carry out cross-modal fusion on the three features, and a feature matrix Ffuse with a time sequence is formed; s3, according to the current diagnosis and treatment information of the patient, establishing a complex common disease network by adopting an advantage ratio RR method; and S4, embedding the feature matrix Ffuse with the time sequence into the complex common disease network by adopting a graph embedding technology, and according to a time sequence data updating rule, forming a cross-modal time sequence common disease network.

The invention relates to the technical field of industrial communication networks, in particular to a method for quickly forwarding messages from a PCIE (Peripheral Component Interface Express) interface to a WIA (Wireless Interface Architecture) interface, which comprises the following steps of: simultaneously receiving PCIE messages and WIA-FA protocol messages through a hardware logic circuit; analyzing the message in a data link layer to obtain an analysis result comprising data load and address information; packaging the analysis result according to a pre-configuration rule to generate a data frame conforming to a target protocol; and constructing a sending descriptor for the generated data frame, and sending the data frame from a corresponding interface through a direct memory access mechanism. According to the method, parallel analysis and protocol conversion of messages are achieved through a hardware logic circuit, fast address conversion is achieved through a pre-configured address mapping table, and zero-copy data transmission is achieved through a direct memory access mechanism. According to the invention, software protocol stack processing is replaced by a hardware processing flow, so that communication delay and CPU resource consumption are effectively reduced.

The invention discloses a task unloading and resource allocation collaborative optimization method for an industrial hybrid network, and belongs to the technical field of industrial communication networks. The method comprises the following steps of: firstly, acquiring multi-dimensional attributes of an industrial field terminal task and a network node state in all directions through intelligent super-sensitive sensing and quantum-level monitoring means; thirdly, constructing a dual-objective function joint optimization model dominated by task total execution time delay minimization and assisted by resource utilization rate maximization, integrating factors such as local queuing time delay and transmission and calculation time delay under the influence of a quantum state, and setting strict constraint conditions; then, an improved quantum heuristic distributed co-evolution algorithm is adopted for solving, and a fine operation and self-adaptive quantum dynamic parameter adjustment mechanism is designed to approach an optimal solution; and finally, allocating tasks and resources according to an optimization result, establishing a continuous monitoring system, and rapidly restarting an optimization process in case of an abnormality.

The present invention relates to the field of network technology and security, and in particular, to a scenario-based self-learning method and system for detecting malicious requests. The method mainly comprises: scenario-based one-click configuration, customized configuration of detection granularity and strategies, and an integrated scoring engine that combines anomaly scores obtained from different detection strategies. In addition, the method also comprises generating configuration feedback suggestions on the basis of traffic self-learning and presenting same to a client in a clear and operable manner. The client can adjust configurations on the basis of the feedback suggestions, generate personalized scenario plans for seamless application, and conduct gray box testing before deployment to ensure the usability of the current scenario-based configuration. The present invention achieves the configuration of flexible and adaptive capabilities for service scenarios by finely categorizing requests for different scenarios, applying different traffic restrictions for different scenarios, and combining different detection functions for different scenarios, so that the security detection capability of scenarios is improved, and missed detections and false positives caused by general configurations are reduced.

The invention discloses a network-on-chip routing method and device, routing nodes, equipment, a medium and a product, and relates to the technical field of network-on-chip, and the method comprises the steps: firstly, screening the routing nodes through employing the input direction of a message, and limiting the selection range of the next-hop routing node; the relative position of the target routing node relative to the current routing node is determined, the routing nodes are further screened according to the relative position, and candidate routing nodes are determined; and determining a next-hop routing node according to the cache capacity and the input direction of the candidate routing node, and transmitting the message to be transmitted to the next-hop routing node to complete routing. The problem that the routing of the network-on-chip is easy to cause network message blockage or link failure of the network-on-chip can be solved. According to the method, an input direction is adopted to limit a selection range of a next-hop routing node, a relative position is adopted to determine candidate routing nodes, the next-hop routing node is determined in the candidate routing nodes based on cache capacity, and self-adaptive deadlock-free routing is realized.

The invention provides a tunnel three-dimensional geologic body modeling method and system based on in-hole camera shooting and while-drilling lithology perception, and relates to the technical field of tunnel and underground engineering and intelligent geological information processing.The method comprises the steps that firstly, a drill hole wall optical image sequence and while-drilling engineering parameters are synchronously collected and preprocessed respectively; then establishing a depth-time mapping relation to realize accurate alignment of the multi-source data in a depth domain; then fracture intelligent identification and parameter extraction are realized through an improved U-Net network, and lithology classification is completed by using an XGBoost model; feature level and decision level fusion is carried out on the fracture features and the lithologic features, and rock mass quality grades are generated; and finally, constructing an integrated geologic model fusing lithologic distribution and a three-dimensional fracture network by adopting an implicit modeling and discrete fracture network technology. According to the method, the problem of splitting of in-hole visual information and while-drilling physical information is solved, automation and precision of geological recognition are improved, and reliable geological information support is provided for tunnel construction safety and support design.

The utilization of drones, mixed reality displays, novel networking techniques, advanced networking techniques (e.g., edge computing), and AI-driven data fusion and visualization. This combination of technologies is employed in tandem to provide advanced data visualization capabilities in mixed reality displays, primarily for usage between multiple devices including mixed reality displays, drones, and other systems that would support networking and data collection capabilities. These capabilities are primarily intended for usage between multiple mixed reality displays, although could be applied to robots, autonomous vehicles, guidance systems, weapon systems, and detection systems.

The present invention belongs to the technical field of communication networks. Disclosed is an adaptive split federated learning method for a distributed AI training service in a 6G computing power network. The method comprises: establishing a 6G computing power network model, and constructing an adaptive split federated learning model for a distributed AI training service; analyzing the delays of each intelligent terminal in a model training stage and a model aggregation stage; constructing a total service delay minimization optimization problem; and using a shortest-path search algorithm to perform joint optimization on model segmentation modes, cooperative node selection modes and multi-domain resource scheduling modes in the model training stage and the model aggregation stage. In the present invention, by means of joint optimization of a model segmentation mode, node selection and multi-dimensional resource allocation in a multi-base-station scenario, the problem of network congestion caused by a single-base-station learning framework processing the aggregation of a large number of intelligent terminal models is alleviated, and the present invention adapts to the heterogeneous computing power of intelligent terminals and the differentiated channel quality in an actual network, thereby improving the efficiency and generalization capability of model training.

Devices, systems, methods, and processes for operating illumination devices on network devices are described herein. Illumination devices, such as light-emitting diodes, on network devices can provide useful status information when being observed by a technician. However, these illumination devices are often not being looked at in large data centers. As such, energy is often wasted on powering these illumination devices. In order to avoid this wasted power usage and increase sustainability, various embodiments described herein turn off illumination devices, such as network indicator lights and overhead lights, as a default state and illuminates them in response to a specific need. This need can be in response to a work order as well as to illuminate a path form an access point of the server room to the target network device that requires work by the network technician. Dynamically engaging illumination devices associated with network devices can reduce overall power usage.